Photochemistry and Photobiology, 2014, 90: 935–940

Measurements of Pilots’ Occupational Solar UV Exposure Adrian Chorley1, Michael Higlett*2, Katarzyna Baczynska2, Robert Hunter3 and Marina Khazova2 1

UK Civil Aviation Authority, Gatwick Airport South, West Sussex, UK Public Health England, Centre for Radiation, Chemical and Environmental Hazards, Didcot, UK 3 BALPA, West Drayton, UK 2

Received 6 January 2014, accepted 4 March 2014, DOI: 10.1111/php.12269

ABSTRACT

(ELVs) take into account the biological effectiveness of the optical radiation in causing harm at different wavelengths, the duration of exposure to the optical radiation and the target tissue. The ELVs represent levels at which ICNIRP considers most of the working population can be repeatedly exposed without suffering any acute adverse health effects and without demonstrated risk of long-term effects. To protect the eye from UV induced cataract, a maximum radiant UV-A (315–400 nm) exposure for the eyes within an 8 h working day should not exceed 10 kJm 2. The ICNIRP guidelines recommend that the UV-A limits be considered as “ceiling values” for the eye and ELVs are directly applicable to exposure of the cornea under worst-case conditions of normal incidence (7). Previous research investigating civilian pilot exposure to UVR and Blue Light hazards is limited (8,9). Diffey and Roscoe (8) measured erythema-weighted radiant exposure during flights using polysulfone film badges worn by pilots on the epaulette nearest to the side window. Recordings were taken from the captain and first officer on 12 flights, including long and short-haul on a wide variety of routes worldwide; the total exposure during flight was then measured from the badge. The results showed that maximum exposures did not exceed 0.019 MED h 1 and were significantly less than the doses on unshaded horizontal surface at ground level: 2.3 MED h 1 in Adelaide (35° S, March) or 0.73 MED h 1 in Newcastle (55° N, partial clouds, June). It was concluded that annual occupational exposure of four MEDs is negligible and civilian aircraft offer virtually complete protection from biologically damaging UVR. As spectral sensitivity of polysulfone dosemeters is restricted to the wavelengths shorter than 330 nm (10), they are not suitable for UV-A measurements; radiant exposure from solar radiation filtered by the aircraft windscreen may be underestimated if measurements do not include the contribution from the UV-A component. Furthermore, polysulfone films register total radiant exposure and do not contain spectral information or dose rate. These data cannot be correlated to flight log or used for providing evidence-based guidance on eye protection. Roscoe and Diffey (9) also used a sensor comprising an internally baffled barrel with 2.5° field of view and blue light filter, 370–520 nm, to measure blue light in a cockpit. During a 2 h 15 min flight from Gatwick to Malaga, a wide variation in radiance was found depending on the direction of flight; the authors concluded that the blue light hazard was similar to that at ground level pointing up at a clear sky. To carry out detailed assessments of pilots’ exposure under different flight conditions, time-stamped spectral data are needed.

It is known that ultraviolet radiation (UVR) increases by 10– 12% every 1000 m altitude; UVR at the 10 000 m of typical cruise altitude for commercial aircraft may be 2–3 times higher than at ground level. Information on the levels of solar UV exposures is essential for the assessment of the occupational risk of pilots developing sun-related eye disorders and skin cancers. The aim of the study was to investigate how UV hazard exposures can be measured during flights so that the occupational dose can be ascertained and compared with international guidance. This article describes the development of instrumentation for automated time-stamped spectral measurements which were collected using bespoke automation software. The software enables the advanced acquisition techniques of automated dark signal capture and multiband integration control optimizing the dynamic performance of the spectrometer over the full spectral range. The equipment was successfully tested in a number of aircraft and helicopter flights during 2012–2013 and illustrated in this article on an example of a Gatwick-Alicante flight.

INTRODUCTION It is known that ultraviolet radiation (UVR) increases by 10– 12% every 1000 m altitude (1); UVR at the typical cruise altitude of commercial aircraft, 10 000 m, may be 2–3 times higher than at ground level. Ocular exposures may further increase due to reflectance from clouds as water reflects both the direct UVR from the sun as well as the diffuse component from the entire sky. Research addressing the prevalence of UVR related ocular pathology in the professional pilot population is limited and inconclusive (2). Investigation of the eye protection habits of professional pilots (3) shows a wide variation in the use of sunglasses during flights. In addition, there is some evidence of an increased prevalence of melanoma in professional pilots (4). Information on the levels of solar UV exposures is essential for the assessment of the occupational risk of pilots developing sunrelated eye disorders and skin cancers. The International Commission on Non-Ionizing Radiation Protection (ICNIRP) has issued exposure limit guidelines for both UVR and Blue Light hazards (5,6). The Exposure Limit Values *Corresponding author email: [email protected] (Michael Higlett) © 2014 Crown copyright. Photochemistry and Photobiology © 2014 The American Society of Photobiology. This article is published with the permission of the Controller of HMSO and the Queen’s Printer for Scotland and Public Health England.

935

936

Adrian Chorley et al.

This could be related to flight details, such as duration, direction of flight, altitude and cloud cover. Miniature CCD array spectroradiometers are increasingly used in a range of applications where rapidly changing spectral information is required to be captured as a function of time, such as in phototherapy dosimetry (11), solar monitoring (12) or measurement of emissions from sunbeds (13,14). Although CCD array spectroradiometers offer many advantages, they suffer from stray light (15) and variation in characteristics with ambient temperature (16). Guidance on their use for measuring solar UV spectra (17) recommends temperature control and automated dark signal acquisition. The constraints of aircraft cockpits and the operational requirements of commercial flights apply additional restrictions on the use of spectroradiometers for the assessment of pilots’ exposure: 1 The confined space rules out temperature control of the instruments; 2 The equipment should not cause interference with aircraft electronics or compromise flight safety; 3 The input optics and light guides may pose an obstruction to pilots if placed near the face; 4 All instrumentation must be independently powered for the duration of the flight.

Figure 1. Components of automated measurement equipment: (a) HR4000 spectroradiometer, (b) optical fiber, (c) in-line TTL shutter with control box and power supply, (d) shutter battery, (e) CC-3-UV diffuser, (f) palmtop computer, (g) battery.

The aim of this study was to investigate how UV exposures can be measured during flights so that the occupational dose can be ascertained and compared with international guidance. This article describes the design of instrumentation, its limitations and illustrates an example of its successful operation on a flight from Gatwick to Alicante.

MATERIALS AND METHODS At any given time, in the absence of additional filters or screens, the shape of the transmitted solar spectrum remains the same and it is independent of the distance from the windscreen, that is the intensity changes but the relative percentage of individual spectral regions remains the same. In other words, UV-A, erythema and Blue–Light-weighted irradiance follow illuminance, and the illuminance is unambiguously linked to the hazard level. It is, therefore, possible to reconstruct UV-A, erythema and Blue–Light-weighted irradiance from measured illuminance using a broad-band instrument if spectral measurements are simultaneously taken at a different distance from the windscreen (18,19). In order not to compromise safe flight operations, it was proposed that measurements of spectral irradiance be carried out at specified times at a fixed position in the cockpit in close proximity to the front aircraft windscreen. Time synchronized broad-band illuminance measurements would be taken near the pilot’s face representing typical tasks during flight. Using broad-band illuminance data and spectral irradiance from spectral measurements, UV-A, erythema and Blue–Light-weighted irradiance may then be determined from the ratio of illuminance at these locations. Measurement hardware. The accessible solar emission was measured over the spectral range of 280–1100 nm using a miniature CCD array spectroradiometer HR4000 (Ocean Optics Inc, Dunedin, FL), S/N HR4C1877, equipped with 25 lm entrance slit and HC1 grating. It was coupled by a metal jacketed QP600-2-SR/BX optical fiber to a CC-3-UV diffuser, see Fig. 1. For fully autonomous operation of equipment during flight, an in-line INLINE-TTL-S fiber shutter was connected directly to the spectroradiometer by RS232. The shutter enabled a dark measurement to be carried out immediately after every data acquisition. The spectroradiometer and TTL shutter were controlled by automation software installed on an ASUS R2E palmtop (Windows Vista operating system) connected to the spectroradiometer through a single USB computer cable. An XCell Pro battery enabled more than 8 h continuous operation of the palmtop and the HR4000; for longer flights a second battery was available. A YSN-12680 12 VDC battery was used to power the optical shutter.

Figure 2. TR-74Ui Illuminance UV Recorder.

Two miniature TR-74Ui Illuminance UV Recorders (T&D Corp, Japan) shown in Fig. 2 were used to record illuminance data that were time-synchronized with spectral measurements. One unit was at a fixed position (set for automated readings), side-by-side with the input optics of the HR4000; the other was used by a researcher taking manual readings from an additional fold down seat located between the pilots’ seats known as the aircraft “jump seat”. These readings were taken at the level of the pilot’s face to measure ocular exposure when looking straight ahead out of the cockpit and angled down to measure ocular exposure when looking toward the primary aircraft instrumentation. A maximum obtainable illuminance reading in the cockpit was also taken. These illuminance data and spectral irradiance from spectral measurements were then used to calculate UV-A and Erythema-weighted irradiance for the assessment of ocular safety for the flight. An assessment was carried out prior to flight to ensure that the equipment did not interfere with any aircraft systems and that the airline captain (responsible for the safe conduct of the flight) approved of the positioning and securing of the equipment for flight. Accuracy of wavelength calibration of the HR4000 was verified before and after each flight by a low pressure Hg pen-ray lamp; Fraunhofer lines were used for confirmation of wavelength stability of in-flight data. The system was calibrated and its performance was monitored over the duration of this study in a laboratory controlled environment using 1000 W tungsten-halogen lamps, calibrated for spectral irradiance to the Physikalisch-Technische Bundensanstalt (PTB) traceable reference standards. For the analysis of the in-flight exposures and to minimize contribution of stray light, the spectroradiometer was additionally calibrated to the solar spectral irradiance at solar noon on a clear day during mid summer, using a scanning double-grating monochromator D3 180 (Jobin Yvon, Longjumeau, France) as a reference instrument. It is important for solar measurements that the input optic is able to collect radiation at different angles. The angular response of the CC-3-

Photochemistry and Photobiology, 2014, 90 UV diffuser was measured using a collimated 100 W tungsten-halogen lamp at angles ranging from 80° to +80° in 5° intervals. For the incident angles (30°), the input optic of the spectral system matches the cosine response within 5% and is constant with wavelength. At 40°, the input optic throughput is approximately 9% lower. The angular response of TR-74Ui Illuminance UV Recorder is similar; the cosine response error is 6% at 30° and 8% at 40°. ICNIRP guidance (7) states that for the eye hazard assessment, the detector field of view can be reduced and limited to 80° (40° from the normal). Angles of incidence higher than 40° are restricted by the cockpit structure and so the angular response of the instrumentation is considered to be suitable for this application. It is known that the performance of CCD array spectroradiometers is affected by variations in ambient temperature (15). Constraints of in-flight measurements rule out the possibility of operation of the instrument in a temperature-controlled environment. To evaluate temperature effects, wavelength position, sensitivity and structure of background signal were measured at the range of foreseeable operation temperatures from 10°C to 40°C. Elevated ambient temperature causes blueshift of wavelength position, exceeding 0.5 nm at 40°C. The sensitivity change with temperature between 22 and 35°C is relatively small, within 2–3%, with respect to the sensitivity at 22°C. However, the mean of dark signal and the standard deviation of the dark signal both increase significantly with increasing integration time above 100 ms and with increasing ambient temperature. A sharp increase in dark signal, and, as a result, loss of signal-to-noise ratio at elevated temperatures for this instrument is a major limiting factor of use outside a temperature-controlled environment. A dark measurement taken immediately after measuring the signal mitigated this effect. It was shown that thermal equilibrium of the HR4000, for example instrument internal temperature, lagged behind the change of ambient temperature for up to 30 min. Therefore, monitoring ambient temperature during field measurements with this particular instrument may be highly inaccurate if ambient temperature used for correction of its performance. The instantaneous board temperature which relates to the internal temperature of the HR4000 was a better and more dynamic predictor of spectroradiometer characteristics than ambient temperature when the instrument was used outside of a temperature-controlled environment. In this study, board temperature of the HR4000 was automatically recorded for each acquisition for indication of required temperature correction. Control acquisition software. The Automated Spectrometer Acquisition System (ASAS) has been designed for operation with Ocean Optics CCD array spectroradiometers when measurements are required to be repeated at specific time intervals under variable illumination conditions. The schedule of measurements, for example start, end and interval times between measurements, is set within ASAS software so that measurements run autonomously. Captured data may be analyzed within ASAS; the results are displayed in tabular and graphical formats. The ASAS program works by automatically determining the acquisition time of the current light conditions for the specified spectral range to reach a user-defined target count level. Between scheduled measurements, the equipment continuously takes acquisitions and estimates the integration time for the next scheduled time. Within each scheduled acquisition, up to three spectral regions can be chosen to optimize the signal-to-noise ratio within a narrower spectral range than the full spectral capability of the instrument. The maximum count level measured by the HR4000 in the 280–1100 nm solar spectrum is at approximately 530 nm; the signal measured at 400 nm is 20–30% of the maximum value; at 350 nm the signal is less than 10% of maximum value, whereas background is nearly constant across the whole spectral range. If the full spectral range is measured in a single acquisition, data at wavelengths shorter than 400 nm may be subject to low signal-to-noise ratios. Splitting the full instrument spectral range into segments enables optimization of the signal in each spectral region separately while allowing saturation outside the region of interest. The choice of spectral regions may be dictated by the target biomarker, for example 315–400 nm UV-A for studies of ocular damage and 380–600 nm for retinal phototoxicity or melatonin entrainment. Selected individual spectral regions could then be “stitched” together to obtain the complete spectrum. If the spectral ranges of these three regions partly overlap, it also provides a useful control measure.

937

For this study, the following spectral regions were chosen: 280– 400 nm, 380–500 nm and the complete spectral region of the HR4000 spectroradiometer, 280–1100 nm. When saturation is permitted outside the restricted spectral range, charge from saturated pixels may leak into adjacent pixels. This effect is especially critical in measurements of the short wavelength UV range where variations in signal level are high. Well depth is specific to the CCD array spectroradiometer; the HR4000 used in this study has a well depth of 16 383 counts. To avoid saturation in the target spectral region and signal nonlinearity near saturation level, the measurement spectral range was set wider than the spectral range of interest, for example 280–450 nm acquisition boundaries were set for the 280–400 nm spectral region and the target count level was set at 15 000, ~90% of the maximum counts. The time interval between scheduled measurements can be set from a few seconds to 99 h. The time interval must be greater than the actual time required to capture, read out and save light and background data. The minimum time interval for acquisition of three spectral regions based on the maximum integration time for the HR4000 (10 s) is 3 min. In this study, a time interval of 10 min was set; measurements for future cockpit studies could be taken more frequently, for example for measurements during taking off/landing or fights through fast changing cloud cover. Data were saved as raw spectral data and, if selected, as spectral irradiance and effective spectral irradiance weighted with a specific action spectrum, providing that the instrument was calibrated for spectral irradiance and that background measurements were available. Built-in spectral weighting could be chosen from UV hazard spectral weighting function S(k) (5,20), Erythema spectral weighting function (19), Blue Light hazard spectral weighting function B(k) (6), Retinal Thermal hazard spectral weighting function R(k) (6) and luminous efficiency weighting V(k) (21). For each measurement, the saved data file contains the raw signals for light and dark signals, the calibration, un-weighted and, if spectral weighting is chosen, the effective irradiance. Results are also displayed graphically.

RESULTS AND DISCUSSION The automated measurement system was deployed during a number of aeroplane and helicopter flights. The sample data presented were taken from a flight on 1 March 2013 from Gatwick (51°N, 0.19°W) to Alicante (38°N, 0.56°W) on an Airbus A321. Stitching UV spectral region R3 and the whole spectral range R1 showed good overlap and overall stitching was not required for the majority of timed acquisitions. The board temperature of the HR4000 during this flight corresponded to the variation in ambient temperature within 22–29°C where sensitivity change is relatively small (2–3%) and temperature correction was not applied for this flight data. At cruise altitude, UV-A irradiance measured at a fixed position on the aircraft windscreen increased by factor of almost 50 compared with the measurement at ground level at departure in the morning, and by a factor of 6–7 compared with ground level at the destination in the early afternoon, see Fig. 3a. Erythema-weighted irradiance varied from negligible (